The discussion on the merit and feasibility of a possible application of Behavioural Economics in Competition Law and Policy has been fierce, particularly so in the context of US Antitrust Law. The fundamental assumption of Behavioural Economics lies in the recognition that human decision making is vulnerable and subject to biases. So-called neo-classical economic models that have hitherto informed competition law and policy (Chicago School, post-Chicago School, Harvard School) have relied on the existence of a ‘homo economicus’ i.e an ‘[economically] rational’ market participant (be it consumer or firm) who is driven by the goal of constant profit maximisation, appropriates all available information and processes it rationally, as well as exhibits perfect willpower and self-interest.
Behavioural Economics (BE) in an effort to provide a more realistic depiction of human [economic] conduct, introduces the concepts of heuristics (shortcuts) and biases. The latter include e.g. ‘bounded rationality’ (people do not continuously collect and assess relevant information and do not always rely upon it when deciding), bounded ‘willpower’ (knowledge that certain behaviour may run contrary to one’s self-interest does not always affect decision-making) and bounded ‘self-interest’ (profit maximisation does not always prevail over every other social goal(s)). Moreover, BE claims to both be able to detect the competition law-relevant limitations in human decision-making, and to take into account those biases that deviate from perfect rationality in a systematic and predictable manner. This happens through recourse to ‘libertarian paternalism’ (cf. e.g. R. Thaler & C. Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (2008); D. Kahneman, Thinking: Fast and Slow (2011)), advocating an arguably possibly more interventionist approach. Identified biases include e.g. inertia, the overconfidence bias, the ‘endowment effect’, the loss and risk aversion effect, the availability bias, the default bias.
Behavioural Economics has been at the root of much controversy. Genuine or solely apparent dividing lines juxtapose BE on the one side and neo-classical economic models on the other side. It has namely been argued that BE seeks to fully supplant neo-classical models of economic theory; BE is too amorphous and yet in need of a central organising principle so as to be of real use; its suitability to inform Competition Law and policy relates to whether it is applied in a common law rather than a civil law context; there is a connection between an enhanced application of BE and the adoption of a rule of law rather than a rule of reason standard; there is a particular link between the goals of competition law and the application of BE.
In view of the above, it is not surprising that the discussion about BE and competition law has been considered to be of lesser relevance in an EU law context, as the latter would seem to embrace/tolerate BE insights far more than would have been the case on the other side of the Atlantic (cf. N. Petit & N. Neyrinck, ‘Behavioral Economics and Abuse of Dominance: A proposed Alternative Reading of the Article 102 TFEU Case-law’, GCLC Working Paper, 2010).
The most evident examples originate from tying/bundling cases involving exclusionary abuses from dominant undertakings (Art. 102 TFEU):
In the Microsoft (Case COMP/C-3/37.792) case the European Commission departed from the findings to be expected from orthodox economic models. Microsoft had allegedly unlawfully bundled the Windows operating system with a further software application, the Windows Media Player. Requirements to be met for tying/bundling to constitute an abuse violating Art. 102 TFEU, include foreclosure of an equally efficient competitor (the ‘as efficient competitor’ test) as set down in the Commission’s Enforcement Guidelines (paras. 59 et seq.) and coercion (either of a physical nature or of an economic one, i.e. a rebate on the combined price of the products) applied upon the demand-side to buy the tied product (in this case the Windows Media Player).
In Microsoft neither a physical nor an economical coercion was given. The Windows Media Player had been installed free of charge as the default media player in PCs running Windows OS, and purchasers had access to rival media players that could all be subsequently installed and used on the same computer (thus excluding the possibility of a stricto sensu physical coercion). Nevertheless, the Commission (as well as the CFI) affirmed the existence of coercion on behalf of Microsoft. Though the approach is at odds with rational economic thinking – and possibly with the approach the Commission seeks to crystallise in its Guidance on Art. 102 TFEU as well – there is no denying the echoes of behavioural thinking, in this specific context of the purchasers’ inertia bias. The decision referred to the low likelihood of end-users ‘in general’ to use alternative media players given the availability of a similar product coming with the Windows OS.
Further influence by Behavioural Economics can be detected in yet another Microsoft case (COMP/39.530) regarding remedies in a tying/bundling context. In this case the Commission was unhappy with Microsoft tying its internet browser internet Explorer to its Windows OS. The remedy which Microsoft offered involved the appearance of a ‘choice screen’ once the OS first connected to the internet and downloaded updates. The screen presented in random order different choices of browsers, amongst which the Internet Explorer too, for the end-user to choose from. Libertarian paternalism advocated by Behavioural Economics scholars is evident in this context, as the end-user is expected to realise that the OS and the browser are two different products that need not be tied together, but is subsequently presented with a choice to make on their own.
The difference between this approach and the one adopted in the case of the Windows Media Player is striking. The Commission had sought in the latter instance to address the perceived abuse through ‘orthodox’ antitrust means, i.e. by constraining Microsoft to offer a version of its OS stripped of the media player in question. As stated above, Microsoft had not charged for its media player (competing media players available could be downloaded free of charge), hence unsurprisingly PC manufacturers chose to install a version of Windows that included the Windows Media Player – the media player-free version of Windows performed particularly badly commercially thus failing to tackle the issue of disconnecting the two products from each other. It would be of great interest to see whether the Commission will be making use of similar tools within its selection and design of remedies, beyond the specificities/particularities of both cases in question; particularly so with regard to the limitations of the effectiveness of providing more information to consumers as a basis for informed choice. BE research has indicated that a surplus of information may prove detrimental at times, as it can be processed according to biases (or even barely processed at all); the importance of both the amount and the framing of the information provided could be of the essence (cf. OFT, ‘What Does Behavioural Economics mean for Competition Policy?’, March 2010, p. 37) – and so would probably caution.
It has been argued (cf. N. Petit & N. Neyrinck, op. cit., p. 14) that BE buttresses a more expansive interpretation of Refusal to Deal cases in the context of EU law compared to its US counterpart, in conjunction with the handling of the ex ante (dis)incentives of firms to invest into innovation. Explanatory analysis refers to the ‘endowment effect’, i.e. the tendency to demand a higher price for a good in one’s ownership rather than the price one would pay to purchase a similar good. As a consequence regulatory intervention would be warranted by BE, as in the case of essential facilities the holders thereof tend to overestimate its value and hence the magnitude of a possible disincentive, should they be ordered to provide access/deal, while demanding disproportionately high prices from their access seeking competitors. Whilst BE has dealt with supply-side irrationality as well, and biased behaviour may well manifest itself at firm (manager, CEO) level as well, it seems at present uncertain whether it would be possible to generalise a cross-case reading of (dis)incentives in Refusal to Deal cases in EU law by recourse to BE.
Concluding, EU Competition Law substantive assessment framework as well as selected case-law from the Commission’s decisional practice and the Courts’ jurisprudence seem to have [occasionally] embraced a point of view akin to the insights of Behavioural Economics in the context of Art. 102 TFEU and resorted to analysis going beyond the neo-classical economic paradigm.
Limitations [should] apply however:
The primary research [in the area of antitrust analysis] is still at a developing/nascent stage. It is not clear whether substantive assessment – if not intuitive (cf. V. Rose, ‘The Role of Behavioral Economics in Competition Law: A Judicial Perspective’, CPI, Spring 2010, Vol. 6 (1)) at least without mentioning it expressis verbis – deviating from neo-classical thought and inspired by BE would extend to further cases: mirroring recurring criticism regarding shortcomings of BE theory in the context of antitrust law, this could be a punctual rather than overall approach (cf. criticism regarding the lack of an organising principle similar to the economic models inspired by the rational decision-making theories – industrial organisation). After all both BE proponents and their counterpart admit that BE has as yet to deal with what has been perceived – and accordingly criticised – as dependence on market specificity. The argument that contrary to popular misconception BE does not challenge nor seek to supplant orthodox economic thought in toto but rather provides insights where appropriate, may be salient and valid, but does not necessarily – as yet – provide full answers as to whether the debate is of the identical pertinence/significance in the context of EU as in US law.
Furthermore, as at this stage the available ‘arsenal’ provided by BE research has not yet fully been deployed, it is unclear which further topics/areas could be covered/addressed more appropriately by applying BE (cf. C. Camerer et. al., ‘Regulation for Conservatives: Behavioral Economics and the case for “Asymmetric Paternalism”’, Un.Penn.L.Rev. 151 (2003), 1211, 1216, claiming that research in BE has gone beyond an initial stage (which was according to critics ‘just a laundry list of departures from rational choice’) towards ‘scientific consolidation’).
Last but not least, as EU competition law needs to come to terms with the new approach to Art. 102 TFEU, it is still open in which further constellations/scenarios BE may become of relevance. Whether this may still be done in accordance to the current Guidance and the developing case-law remains to be seen.